Russell, E. W., Russell, S. L. K., & Hill, B. D. (2005). The fundamental psychometric status of neuropsychological batteries. Archives of Clinical Neuropsychology, 20, 785-794.
A fundamental requirement for neuropsychological assessment is dependability. Neuropsychological knowledge is dependable only if it has been validated using psychometric methods. Since batteries are used for interpretations, the psychometric validation methods that are acceptable for individual tests must be applied to batteries to produce dependable information. While the standardized battery has been validated, the flexible battery has not. Due to the probability that some tests will be impaired by chance, a flexible battery cannot produce dependable interpretations by selecting or combining test results. Localization and diagnostic assessments are obtained by comparisons. Comparisons require that the tests in a battery are invariant or have equivalent norms along with a common metric. While standardized batteries do meet these criteria, flexible batteries do not. Consequently, clinical judgment applied to a flexible battery cannot provide dependable knowledge beyond that which could be provided by a single validated individual test.
A related article:
McKinzey, R. K., & Ziegler, T. (1999). Challenging a flexible neuropsychological battery under Kelly/Frye: A case study. Behavioral Sciences & the Law, 17(4), 543-551.
The ability of the flexible neuropsychological battery approach to withstand a challenge under California's evidentiary standard, Kelly/Frye, was tested in an actual trial. Despite repeating many criticisms offered by the literature (e.g., no malingering measures, unknown accuracy rates, ignoring statistical limitations, not using age norms), the battery was allowed in "for the weight of the evidence," rather than being excluded as unreliable.
A related court case decision:
Shelby Baxter v Charles & Kelly Temple (Merrimack Co., NH, Superior Court 2005).
In a NH personal injury case, a flexible neuropsychological battery was subjected to a Daubert evidentiary challenge. The judge excluded it, citing the inherent untestability and therefore unknown error rates. He also cited the expert's violation of standardized administration, which invalidated the results. (Editor's abstract)